Search

Glaucoma is the leading cause of irreversible blindness and disability worldwide. Nevertheless, the majority of patients do not know they have the disease and detection of glaucoma progression using standard technology remains a challenge in clinical practice. Artificial intelligence (AI) is an expanding field that offers the potential to improve diagnosis and screening for glaucoma with minimal reliance on human input. Deep learning (DL) algorithms have risen to the forefront of AI by providing nearly human-level performance, at times exceeding the performance of humans for detection of glaucoma on structural and functional tests. A succinct summary of present studies and challenges to be addressed in this field is needed. Following PRISMA guidelines, we conducted a systematic review of studies that applied DL methods for detection of glaucoma using color fundus photographs, optical coherence tomography (OCT), or standard automated perimetry (SAP). In this review article we describe recent advances in DL as applied to the diagnosis of glaucoma and glaucoma progression for application in screening and clinical settings, as well as the challenges that remain when applying this novel technique in glaucoma.

Purpose: To investigate rates of structural and functional change in a large clinical population of glaucoma and glaucoma suspect patients., Design: Retrospective cohort., Methods: Twenty-nine thousand five hundred forty-eight spectral-domain optical coherence tomography (OCT) and 19,812 standard automated perimetry (SAP) tests from 6138 eyes of 3669 patients with ≥6 months of follow-up, 2 good quality spectral-domain OCT peripapillary retinal nerve fiber layer scans, and 2 reliable SAP tests were included. Data were extracted from the Duke Glaucoma Registry, a large database of electronic health records of patients from the Duke Eye Center and satellite clinics. Rates of change for the 2 metrics were obtained using linear mixed models, categorized according to pre-established cutoffs, and analyzed according to the severity of the disease., Results: Average rates of change were -0.73 ± 0.80 μm per year for global retinal nerve fiber layer thickness and -0.09 ± 0.36 dB per year for SAP mean deviation. More than one quarter (26.6%) of eyes were classified as having at least a moderate rate of change by spectral-domain OCT vs 9.1% by SAP (P < .001). In eyes with severe disease, 31.6% were classified as progressing at moderate or faster rates by SAP vs 26.5% by spectral-domain OCT (P = .055). Most eyes classified as fast by spectral-domain OCT were classified as slow by SAP and vice versa., Conclusion: Although most patients under routine care had slow rates of progression, a substantial proportion had rates that could potentially result in major losses if sustained over time. Both structural and functional tests should be used to monitor glaucoma, and spectral-domain OCT still has a relevant role in detecting fast progressors in advanced disease., (Copyright © 2020. Published by Elsevier Inc.)

Purpose: To investigate the impact of intraocular pressure (IOP) control on rates of change of spectral-domain OCT (SD-OCT) retinal nerve fiber layer (RNFL) thickness in a large clinical population., Design: Retrospective cohort study., Participants: A total of 85 835 IOP measurements and 60 223 SD-OCT tests from 14 790 eyes of 7844 patients., Methods: Data were extracted from the Duke Glaucoma Registry, a large database of electronic medical records of patients with glaucoma and suspected disease followed over time at the Duke Eye Center and satellite clinics. All records from patients with a minimum of 6 months of follow-up and at least 2 good-quality SD-OCT scans and 2 clinical visits with Goldmann applanation tonometry were included. Eyes were categorized according to the frequency of visits with IOP below cutoffs of 21 mmHg, 18 mmHg, and 15 mmHg over time. Rates of change for global RNFL thickness were obtained using linear mixed models and classified as slow if change was slower than -1.0 μm/year; moderate if between -1.0 and -2.0 μm/year; and fast if faster than -2.0 μm/year. Multivariable models were adjusted for age, gender, race, diagnosis, central corneal thickness, follow-up time, and baseline disease severity., Main Outcome Measures: Rates of change in SD-OCT RNFL thickness according to levels of IOP control., Results: Eyes had a mean follow-up of 3.5±1.9 years. Average rate of change in RNFL thickness was -0.68±0.59 μm/year. Each 1 mmHg higher mean IOP was associated with 0.05 μm/year faster RNFL loss (P < 0.001) after adjustment for potentially confounding variables. For eyes that had fast progression, 41% of them had IOP <21 mmHg in all visits during follow-up, whereas 20% of them had all visits with IOP <18 mmHg, but only 9% of them had all visits with IOP <15 mmHg., Conclusions: Intraocular pressure was significantly associated with rates of progressive RNFL loss in a large clinical population. Eyes with stricter IOP control over follow-up visits had a smaller chance of exhibiting fast deterioration. Our findings may assist clinicians in establishing target pressures in clinical practice., (Copyright © 2020 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

Purpose: To determine the prevalence of artifacts on segmented spectral-domain optical coherence tomography (SDOCT) images and assess their impact on the interpretation of glaucomatous progression in the retinal nerve fiber layer (RNFL) profile and macular thickness map., Design: Retrospective reliability analysis., Methods: Retrospective review of glaucoma and glaucoma suspect eyes imaged with SDOCT during a 1-month period. All cases had at least 4 sets of RNFL and macular images at 6-month intervals. SDOCT raw B-scans were examined to determine true progression and whether artifacts impacted the original interpretation of progression based on auto-segmented change maps. The co-prevalence of artifacts in the RNFL and macula was assessed, as well as the association of clinical factors with the likelihood of artifacts., Results: A total of 190 eyes with 760 sets of OCT RNFL and macular scans were included. Fifty percent (96/190) of eyes had artifacts, either in the circumpapillary RNFL (83/190; 43.68%) or the macula (57/190; 30.0%). Epiretinal membrane and vitreomacular traction were the most common artifacts. True progression was present on 39.5% (75/190) of scans overall. Among scans with artifacts, 23.9% (23/96) of artifacts masked true progression (ie, false-negative), 36.5% (35/96) led to an interpretation of false progression (ie, false-positive), and 39.6% (38/96) had no effect on the interpretation of progression. The presence of true progression on the RNFL scan was significantly associated with the presence of true progression on the macular scan (P < .001). Similarly, the presence of artifacts on the RNFL scan was significantly associated with artifacts on the macular scan (P < .001). In multivariable analysis, severe glaucoma, hypertension, and age were significantly associated with the presence of artifacts on RNFL (P < .05)., Conclusions: Artifacts are highly prevalent on both circumpapillary RNFL and macular scans on SDOCT images acquired in a glaucoma clinic. Artifacts can lead to false-positive and false-negative interpretation of progression when using only the auto-segmentation change maps. Thus, careful examination of the raw B-scan images of both the RNFL and macula is critical to identify artifacts and true glaucoma progression., (Copyright © 2020 Elsevier Inc. All rights reserved.)

Purpose: The rule of 5 is a simple rule for detecting retinal nerve fiber layer (RNFL) change on spectral-domain OCT (SD-OCT), in which a loss of 5 μm of global RNFL on a follow-up test is considered evidence of significant change when compared with the baseline. The rule is based on short-term test-retest variability of SD-OCT and is often used in clinical practice. The purpose of this study was to compare the rule of 5 with trend-based analysis of global RNFL thickness over time for detecting glaucomatous progression., Design: Prospective cohort., Participants: A total of 300 eyes of 210 glaucoma subjects followed for an average of 5.4±1.5 years with a median of 11 (interquartile range, 7-14) visits., Methods: Trend-based analysis was performed by ordinary least-squares (OLS) linear regression of global RNFL thickness over time. For estimation of specificity, false-positives were obtained by assessing for progression on series of randomly permutated follow-up visits for each eye, which removes any systematic trend over time. The specificity of trend-based analysis was matched to that of the rule of 5 to allow meaningful comparison of the "hit rate," or the proportion of glaucoma eyes categorized as progressing at each time point, using the original sequence of visits., Main Outcome Measures: Comparison between hit rates of trend-analysis versus rule of 5 at matched specificity., Results: After 5 years, the simple rule of 5 identified 37.5% of eyes as progressing at a specificity of 81.1%. At the same specificity, the hit rate for trend-based analysis was significantly greater than that of the rule of 5 (62.9% vs. 37.5%; P < 0.001). If the rule of 5 was required to be repeatable on a consecutive test, specificity improved to 93.4%, but hit rate decreased to 21.0%. At this higher specificity, trend-based analysis still had a significantly greater hit rate than the rule of 5 (47.4% vs. 21.0%, respectively; P < 0.001)., Conclusions: Trend-based analysis was superior to the simple rule of 5 for identifying progression in glaucoma eyes and should be preferred as a method for longitudinal assessment of global SD-OCT RNFL change over time., (Copyright © 2020 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

Because of recent advances in computing technology and the availability of large datasets, deep learning has risen to the forefront of artificial intelligence, with performances that often equal, or sometimes even exceed, those of human subjects on a variety of tasks, especially those related to image classification and pattern recognition. As one of the medical fields that is highly dependent on ancillary imaging tests, ophthalmology has been in a prime position to witness the application of deep learning algorithms that can help analyze the vast amount of data coming from those tests. In particular, glaucoma stands as one of the conditions where application of deep learning algorithms could potentially lead to better use of the vast amount of information coming from structural and functional tests evaluating the optic nerve and macula. The purpose of this article is to critically review recent applications of deep learning models in glaucoma, discussing their advantages but also focusing on the challenges inherent to the development of such models for screening, diagnosis and detection of progression. After a brief general overview of deep learning and how it compares to traditional machine learning classifiers, we discuss issues related to the training and validation of deep learning models and how they specifically apply to glaucoma. We then discuss specific scenarios where deep learning has been proposed for use in glaucoma, such as screening with fundus photography, and diagnosis and detection of glaucoma progression with optical coherence tomography and standard automated perimetry., Translational Relevance: Deep learning algorithms have the potential to significantly improve diagnostic capabilities in glaucoma, but their application in clinical practice requires careful validation, with consideration of the target population, the reference standards used to build the models, and potential sources of bias., Competing Interests: Disclosure: A.C. Thompson, None; A.A. Jammal, None; F.A. Medeiros, Aeri Pharmaceuticals (C); Allergan (C, F), Annexon (C); Biogen (C); Carl Zeiss Meditec (C, F), Galimedix (C); Google Inc. (F); Heidelberg Engineering (F), IDx (C); nGoggle Inc. (P), Novartis (F); Stealth Biotherapeutics (C); Reichert (C, F), (Copyright 2020 The Authors.)

Purpose: To assess short- and long-term variability on standard automated perimetry (SAP) and spectral domain optical coherence tomography (SD-OCT) in glaucoma., Design: Prospective cohort., Methods: Ordinary least squares linear regression of SAP mean deviation (MD) and SD-OCT global retinal nerve fiber layer (RNFL) thickness were fitted over time for sequential tests conducted within 5 weeks (short-term testing) and annually (long-term testing). Residuals were obtained by subtracting the predicted and observed values, and each patient's standard deviation (SD) of the residuals was used as a measure of variability. Wilcoxon signed-rank test was performed to test the hypothesis of equality between short- and long-term variability., Results: A total of 43 eyes of 43 glaucoma subjects were included. Subjects had a mean 4.5 ± 0.8 SAP and OCT tests for short-term variability assessment. For long-term variability, the same number of tests were performed and results annually collected over an average of 4.0 ± 0.8 years. The average SD of the residuals was significantly higher in the long-term than in the short-term period for both tests: 1.05 ± 0.70 dB vs. 0.61 ± 0.34 dB, respectively (P < 0.001) for SAP MD and 1.95 ± 1.86 μm vs. 0.81 ± 0.56 μm, respectively (P < 0.001) for SD-OCT RNFL thickness., Conclusions: Long-term variability was higher than short-term variability on SD-OCT and SAP. Because current event-based algorithms for detection of glaucoma progression on SAP and SD-OCT have relied on short-term variability data to establish their normative databases, these algorithms may be underestimating the variability in the long-term and thus may overestimate progression over time., (Copyright © 2019 Elsevier Inc. All rights reserved.)

This study describes a segmentation-free deep learning (DL) algorithm for measuring retinal nerve fibre layer (RNFL) thickness on spectral-domain optical coherence tomography (SDOCT). The study included 25,285 B-scans from 1,338 eyes of 706 subjects. Training was done to predict RNFL thickness from raw unsegmented scans using conventional RNFL thickness measurements from good quality images as targets, forcing the DL algorithm to learn its own representation of RNFL. The algorithm was tested in three different sets: (1) images without segmentation errors or artefacts, (2) low-quality images with segmentation errors, and (3) images with other artefacts. In test set 1, segmentation-free RNFL predictions were highly correlated with conventional RNFL thickness (r = 0.983, P < 0.001). In test set 2, segmentation-free predictions had higher correlation with the best available estimate (tests with good quality taken in the same date) compared to those from the conventional algorithm (r = 0.972 vs. r = 0.829, respectively; P < 0.001). Segmentation-free predictions were also better in test set 3 (r = 0.940 vs. r = 0.640, P < 0.001). In conclusion, a novel segmentation-free algorithm to extract RNFL thickness performed similarly to the conventional method in good quality images and better in images with errors or other artefacts.

Purpose: To evaluate whether loss of 5 μm or more in global retinal nerve fiber layer (RNFL) thickness on spectral-domain (SD) between 2 consecutive visits is specific for glaucoma progression., Design: Prospective cohort., Participants: Ninety-two eyes of 49 control participants and 300 eyes of 210 glaucoma patients., Methods: Patients completed at least 5 standard automated perimetry and SD OCT examinations at 6-month intervals over at least 2 years. Eyes were categorized as progressing from glaucoma if the average RNFL declined by 5 μm or more between 2 consecutive visits. The false-positive proportion was estimated by 2 methods: (1) 5-μm or more loss in control participants and (2) 5-μm or more gain in glaucoma. The false-positive proportion was subtracted from the cumulative proportion of eyes categorized with glaucoma progression to estimate the true progression prevalence., Main Outcome Measures: False-positive and true progression prevalence of patients with glaucoma detected as progressing on SD OCT., Results: After 5 years of semiannual testing, the cumulative proportion of false-positive results based on 5-μm or more RNFL losses between visits was 24.8% in the control participants. Although 40.6% of glaucoma eyes were diagnosed with progression at 5 years, only 15.8% would have been considered to show true progression based on the expected false-positive ratio from the control participants (i.e., 40.6%-24.8%). The cumulative proportion of an intervisit gain of 5 μm or more at 5 years was 27.4% in glaucoma eyes, suggesting that only 13.2% of eyes with glaucoma truly had progressed (i.e., 40.6%-27.4%)., Conclusions: Loss of 5 μm or more in average RNFL thickness between consecutive SD OCT tests is not specific for glaucoma progression. Application of this intervisit rule of 5 can result in a high cumulative proportion of false-positive results over time, which could lead to unnecessary interventions in patients whose disease is stable. More specific diagnostic criteria are needed to help clinicians determine whether patients with glaucoma are progressing so that therapy escalation is both timely and appropriate., (Copyright © 2019 American Academy of Ophthalmology. Published by Elsevier Inc. All rights reserved.)

In this study we developed a deep learning (DL) algorithm that detects errors in retinal never fibre layer (RNFL) segmentation on spectral-domain optical coherence tomography (SDOCT) B-scans using human grades as the reference standard. A dataset of 25,250 SDOCT B-scans reviewed for segmentation errors by human graders was randomly divided into validation plus training (50%) and test (50%) sets. The performance of the DL algorithm was evaluated in the test sample by outputting a probability of having a segmentation error for each B-scan. The ability of the algorithm to detect segmentation errors was evaluated with the area under the receiver operating characteristic (ROC) curve. Mean DL probabilities of segmentation error in the test sample were 0.90 ± 0.17 vs. 0.12 ± 0.22 (P < 0.001) for scans with and without segmentation errors, respectively. The DL algorithm had an area under the ROC curve of 0.979 (95% CI: 0.974 to 0.984) and an overall accuracy of 92.4%. For the B-scans with severe segmentation errors in the test sample, the DL algorithm was 98.9% sensitive. This algorithm can help clinicians and researchers review images for artifacts in SDOCT tests in a timely manner and avoid inaccurate diagnostic interpretations.

Purpose: The purpose of this study was to investigate factors associated with a second exposure of a glaucoma drainage device (GDD) following repair of an initial GDD exposure., Materials and Methods: This IRB-approved retrospective cohort study examined the incidence of a second exposure of a GDD following initial repair for exposure. Logistic regression was performed to assess the relationship between demographic and clinical characteristics and a second exposure of the GDD. Kaplan-Meier survival curves were plotted and Cox regression was performed to examine factors impacting the time to a second GDD exposure., Results: Ninety-four eyes of subjects that underwent initial revision for GDD exposure were reviewed. Approximately 44% (N=41/94) of subjects underwent surgical revision for a second exposure. Factors associated with reexposure in multivariate logistic regression included caucasian race (odds ratio, 2.99; P=0.02) and use of a nonscleral patch graft (odds ratio, 2.93; P=0.019). Time from revision of the initial exposure to reexposure was significantly shorter for those with a nonscleral patch graft (hazard ratio, 2.23; P=0.01) and caucasian race (hazard ratio, 2.08; P=0.04)., Conclusions: Caucasian race and use of a nonscleral patch graft during revision surgery was associated with a higher risk of experiencing a sooner reexposure of the GDD following revision of an initial exposure. Future studies should examine whether particular graft materials increase the risk of GDD reexposure.

Introduction: To evaluate the agreement between trend-based analysis and qualitative assessment of the retinal nerve fiber layer (RNFL) thickness for glaucomatous progression on spectral-domain optical coherence tomography (SDOCT). Methods: Retrospective review of 190 eyes from 103 patients with glaucoma or suspected glaucoma that underwent SDOCT imaging during four consecutive clinic visits. Trend-based progression was characterized by a significantly negative slope. Progression by qualitative analysis was determined by review of raw SDOCT B-scans. Results: The slope was significantly greater in those with progression than without progression for both trend-based and qualitative analysis (p < 0.001). However, the qualitative grading classified a significantly greater proportion of eyes as progressing compared to trend-based analysis in both the superotemporal (ST) (23.2% vs. 10.5%, p = 0.001) and inferotemporal (IT) RNFL (27.4% vs 8.4%, p < 0.001). The trend-based and qualitative classifications of progression showed poor agreement in both the ST (kappa = 0.0135) and IT RNFL (kappa = 0.1222). The agreement between trend-based and qualitative analysis was lower for eyes with artifacts (ST = 58.11%; IT = 68.7%) than those without artifacts (ST = 80.2%; IT = 74.8%). Moreover, among eyes with artifacts, there was no significant difference in slope between those qualitatively categorized as progressing versus not progressing (p > 0.05). Conclusions: Poor agreement was found between a trend-based and qualitative analysis of change in RNFL on SDOCT. Careful qualitative review of SDOCT imaging may identify specific areas of glaucoma progression not captured by trend-based methods, especially in the presence of artifacts. Such an approach may also prove useful for detecting glaucoma progression in a clinical setting when there are few data points available. [ABSTRACT FROM AUTHOR]

Several prior studies have described the co-prevalence of peripapillary RNFL retinoschisis in eyes with glaucoma [[2]-[5]], which may impact approximately 6% of glaucoma patients according to two recent cohort studies [[3]]. Thus, when reviewing serial SD-OCT in patients with glaucoma, it is critical to neither overestimate the RNFL thickness in eyes with retinoschisis, nor erroneously attribute the resolution of such retinoschisis to glaucomatous progression. Van der Schoot et al. detected focal peripapillary RNFL retinoschisis in 7 of 117 glaucomatous eyes and 0 of 91 healthy control eyes [[4]]. [Extracted from the article]